UNLEASHcast: Evolution and application of AI with Nikolas Kairinos

Welcome to the first of many guest interviews in the UNLEASHcast audio library – on this occasion with Soffos.ai founder Nikolas Kairinos. We talk about the changing face of artificial intelligence (AI) and the many irresistible business opportunities it brings.

Check out the the audio above, or the full transcript beneath…

+++

UNLEASH: Thanks so much for talking to UNLEASH today, we’re going to be talking about AI and the myriad ways in which it’s used applied and is essential for modern business – but a bit of background about you. You’ve been involved in AI for four decades. That’s a seriously long time to be involved in artificial intelligence; I guess a lot of people might not realize that it’s a technology that’s been around for so long. So what’s changed in your experience?

Nikolas Kairinos: I got started in AI as a kid, really; I started programming at the age of 11. And I became enamored with this concept of AI and robotics, and sort of taught myself everything and anything I could, experimenting and developing basic AI applications over the years. Obviously, with the various AI winters, for large parts it was just a hobby, but I was doing my own research playing around.

I think what hasn’t changed is a lot of the false expectations, right? Unrealistic expectations, often hyped by the media. And that’s what caused the AI winters and the lack of funding and development back at the time, because you’d see some amazing things happening. And then those would get hyped up to, okay, well, if we could do that, then that we just project linearly, and all these other amazing human-like things will be able to be done by machines. And when that didn’t materialize, people lost interest.

But the diehard researchers and practitioners like myself just kept on prodding. And what’s happened is largely thanks to Geoff Hinton and his colleagues in that the genesis of machine learning and deep learning, essentially, is both a blessing and a curse. It’s amazing, in that it gives us a lot of phenomenal technology.

But it’s a curse in that it’s become the total focus of most people in AI, they’ve forgotten a lot about a lot of the good old fashioned AI, that has its roots in symbolic AI, that considers things like logic, inputs and outputs that humans can read and understand and that sort of grounded AI, which would have taken longer to get us to where we need to get to.

There’s a lot of hype as to what it can do. I mean, Geoffrey Hinton and others say that it’ll be able to do everything eventually. I don’t buy that because things are not probabilistic and statistical. Not everything is completely mathematical. It ignores semantics, and logic, and reason, and consciousness, and all these things that are truly human traits and make us intelligent. So I could go on and on about all the things that have changed.

But what I’ve started to see happen now, and it’s something that we do a lot of, is starting to merge the sort of neuro neural network probabilistic approaches that you see, and the original form of AI, symbolic AI and kind of merge those two together into the best of both and having them work together to achieve something that is akin to artificial general intelligence, which is the holy grail, which we may or may not see in our lifetime.

UNLEASH: [AI] seems to me that it’s crystallizing for lots of people now. And more and more businesses are adopting it in different ways. As the technology diverges, and is applied in more different directions, how do you think AI can help HR systems, HR technology, other aspects of that world?

NK: In many, many ways. I mean, you’re seeing a lot of applications come up from the recruitment process alone, you know? We’re a small company and when we run a job ad, we are inundated with applications, right? So it’s not too difficult to use.

Very basic AI can do a lot of the pre-screening, and filtering and even ask questions and do some of the pre-interview processes, which will save us a tremendous amount of time and help us identify traits in new applicants that match the traits and attributes of our most successful employees. So do more of what works.

What I don’t want to see happening, and I see companies talking about it is having the AI have the ultimate decision; it should be a tool that we use to help crunch the numbers and filter out obviously bad matches, but based on non-biased criteria.

A big area in our world, which is related to HR, is learning and development. In the corporate setting, AI is going to play a massive role there, so one of the biggest problems we are [looking] to solve is this idea that information is flowing faster than ever in history. Right?

It used to be, 100 years ago, that you could spend a decade researching something, spend another five years writing a textbook, getting it peer-reviewed, published, and then starting to teach it to your students over a 20-year gap. You can’t do that today. Because by the time you publish anything, much of it is obsolete, because there’s this new flow of knowledge coming our way. And to make matters even more complex, a lot of the knowledge is not factual. It’s not scientifically reviewed and checked and validated. So you’ve got to filter out the noise, filter out that disinformation.

So that’s a massive area, taking all this new knowledge and making it available real time to employees. So whether I’m a researcher working in a lab, whether I’m an accountant, it doesn’t make a difference what role I play in the company, if I have information at my fingertips, without having to have ‘total recall’, and perhaps even not even knowing what I need to be asking.

In other words, [AI] starts connecting the dots for me saying, well, you’re looking at this, this and this, there’s a whole big missing piece you haven’t looked at here. So I suggest you go learn this adjacent knowledge because it’s going to enable you to do your job better – giving employees knowledge when they need it on the fly. And then, improving training is an area that I’m super excited about and one that we’ve devoted years to. We’re just beginning to release products that talk to that and allow employees, as well as students and consumers, to learn better and be more effective.

UNLEASH: You mentioned misinformation and disinformation; do you think that now, people are generally a little more comfortable with the idea of AI? Because there’s the ‘they’re taking our jobs’ idea and all of that. But as it’s been adopted and adapted more, we’re seeing more of a positive view about AI and how it can help people. There’s a lot of possible skeptics, but is that number getting smaller?

NK: My whole philosophy to AI is that it should guide and give us options, but never make decisions that have big implications. Every time I speak at an event or on stage on a panel, inevitably, somebody will ask the question about AI taking our jobs, right?

We know at the moment that it is very difficult to find people because you you’ve got the mass resignation, [and] the new generation just does not really want to work the way we used to work. So there’s going to be a massive shortage of labor. And therefore it’s now flipped around. It’s not ‘is the AI going to take our jobs?’ It’s ‘…we better get our stuff together really quickly, and get the AI up to speed because we’re going to need it to fill in the jobs that nobody wants to do anymore.’

So it’s actually become an imperative. It’s become not a fear that people should have – we should fear that it doesn’t develop fast enough to actually fill in those gaps. Because that’s a real real problem.

UNLEASH:  I don’t think [The Great Resignation] is what I or many other people predicted what would happen from the outcome of the pandemic. Obviously, people are moving, leaving jobs, going to going to pursue their own projects for whatever reason, unacceptable working conditions, etc.

But the idea of work from anywhere; that’s going to be a big, positive part of the future of work, I think. The idea of a geographically dispersed workforce. The need for really good virtual working conditions and video platforms. Where do you think AI is going to play into that? It’s got to play a huge part in the future of work and what comes after the great resignation, right?

NK: It’s going to happen. There’s a massive opportunity for entrepreneurs in the space. First of all, there’s a lot of disconnects that happen, even in this conversation.

If you and I were sitting across the table from each other, there’d be a stronger connection, we’d be feeding more off each other. Magic happens when you put people in a room. We just see it again And again. I don’t have scientific data to back that up. But anecdotally, it’s super clear: AI needs to be developed.

And it’s not just AI, it’s all the related converging technologies, virtual reality, and so on and so forth. Even knowledge management systems – an AI could be listening to this conversation, automatically updating a knowledge management system, and then disseminating that information to whomever else in the organization or in the stakeholder network that needs to hear it. Better video platforms, 3D virtual reality, things like that need to happen. There’s a lot of hype around the metaverse; I’d very much like to see that applied within the learning and corporate environments as well.

So, there’s people who want to work remote, and not only that, like in our industry, it’s so hard to find people with the skills that we need to do what we’re doing, because as a deep tech company, there’s maybe half a dozen people at any point that can do a specific thing, and the idea of having them all in one place just doesn’t make sense.

We don’t care where you situated as long as you can do the job. So a lot of new technologies are coming that will enable that and improve on that. So watch this space – there’s there’s a lot coming.

UNLEASH: Well, Nicholas, thanks for talking to us. So much opportunity here, I think. Great to talk to you.

NK: Absolutely. Thanks for having me.